Alek wrote:
>Today we got some issues calculating the future dates.
...
>27.09.2010 03:02:25
>> Invalid local time for date in time zone: Europe/Helsinki

The problem is exactly what it says.  Adding six months to your input
date, you're trying to work with the local time 2011-03-27T03:02:25,
which *doesn't exist* in Helsinki.  At 2011-03-27T00:59:59Z Helsinki will
be on UT+2h (EST), with a local time of 2011-03-27T02:59:59.  But then
at 2011-03-27T01:00:00Z it will move to UT+3h (EEST), with a local time
of 2011-03-27T04:00:00.  There's a related problem at the end of October,
when certain local times of day occur twice in one day.

This kind of situation is unavoidable, in the general case, when working
with civil timezones.  You can avoid it for a particular timezone by only
using times-of-day that don't coincide with offset changes.  Helsinki, for
example, never switches at local noon.  Beware that some other zones do,
and Helsinki might start doing so at the whim of the Finnish Parliament.

-zefram

Reply via email to